The company in the previous example decides to test an 8-run full factorial design with 3 factors
Question:
The company in the previous example decides to test an 8-run full factorial design with 3 factors at 2 levels. A brainstorming session conducted with the operator, supervisor, and engineer resulted in the finished part being measured at 4 places. ( Two measures were taken on each end 180 degrees from each other.) The designed and resultant response values were:
Run | RPM | Feed Rate | Tool Radius | Surface Finish |
1 | 588 | .004 | 1/64 | 50, 50, 55, 50 |
2 | 588 | .004 | 1/32 | 145, 150, 100, 110 |
3 | 588 | .008 | 1/64 | 160, 160, 155, 160 |
4 | 588 | .008 | 1/32 | 180, 200, 190, 195 |
5 | 1182 | .004 | 1/64 | 60, 60, 60, 55 |
6 | 1182 | .004 | 1/32 | 25, 35, 35, 30 |
7 | 1182 | .008 | 1/64 | 160, 160, 160, 160 |
8 | 1182 | .008 | 1/32 | 80, 70, 70, 80 |
a) Calculate the averages, effects, and half effects for all factors and interactions. Plot the average values for each factor.
b) Generate a Pareto chart of the absolute value of each half effect. Which effects appear to be most important.
c) Generate an interaction plot for any interactions with large effects.
d) What are the best settings for the factors to achieve this objective?
Fundamentals of Corporate Finance
ISBN: 978-0077861629
8th Edition
Authors: Stephen A. Ross, Randolph W. Westerfield, Bradford D.Jordan